AMD: Considering to Break Up the Company in two Seperate Businesses to Compete with Intel

AMD hasn't been doing well since around 2009. it's been 6 years, and AMD hasn't brought a ground-breaking CPU that rightfully competed with Intel in this time-frame. people seem to forget that AMD used to kick Intel's Ass in performance. and truly innovated the CPU market.

For those who don't already know AMD were the FIRST to bring these to market.

  • 1GHz 64-BIT Processor in 2003 (Athlon 64)
  • Dual-Core Processor in 2004 (Athlon X2)
  • Quad-Core Processor in 2007 (Phenom I)
  • Hexa-Core Processor in 2010 (Phenom II X6)
  • Octa-Core Processor 2011 (FX-8150)
  • 5Ghz Processor in 2014 (FX-9590)
  • Develop a Graphics card with High Bandwidth Memory (2015)

To keep this list short. the reason Intel is primarily successful today is due to AMD's Innovation.

but back to the point, AMD hasn't done well in years. and it's not even an argument anymore, AMD purchasing ATI back in 2006 saved AMDs skin. now CEO LIsa Su is taking into consideration ALL options to make AMD the once successful company it once was. even breaking up the company into subsidiaries and or desperate businesses to compete with Intel. even as going as far as to breaking apart the Graphics division and Server division. there may be a possibility of the "ATI" Name may be returning in the future. Nothing is set in stone, but AMD's future could either be Bright or Dark.

Thoughts? I don't know how to properly comment on AMD's situation anymore.


1 Like

Well they've got a nice chunk of cash in reserve still, and the graphics market is still pretty strong. Their embedded market is awesome, with the Jaguar APU dominating consoles. Just pointing that out you see two sides of AMD that can become complete subdivisions

I think they can split the company and still be "AMD," and really that's a great idea. Splitting divisions so that one fallback doesn't halt innovation for every facet and industry they have a hand in. I believe this is something they should absolutely consider while they still have cash in hand and interest from investors. (However little that may be anymore.)

If they decide to go with it I can only wish them the best in succeeding. If not for the competition in the industry, then for the people who call AMD their job.

hmm, doubtful.
Especially after they acquired few memory companies before developing their HBM.

I would love to ATi again, as I loved the way they were, but i don't see how those separate companies could keep up alone. Basically not an option, they need to spread out and chew small companies with their patents to survive.

while AMD might have advanced some spices in cpu industry, it wasn't leader. Intel was better option even though intel had bad time with Pentium 4, it didn't stop intel because of supreme marketing. Intel was still able to sell more than AMD. MMX, SSE simply nailed it to wall, and amd brought MMX from intel and was still killed off because of windows, they didn't call it on amd cpu's which is first big fail as dev's didn't use AMD MMX as they were calling on intel mmx in their software. From there its all down for amd.

I don't see it happening, its good to speak about it because its going to lower shares and its highly possible with fury x their shares going to rise up ~ time to sell sell sell their stock for high price ~

Still i would love to see ATi logo back.

amd stock is going up most likely it'll going to reach $3.7

You can add HBM to that list

1 Like

You could term AMD's move as a 'de-merger' to enable them to offset or alleviate business-risk by splitting-up viable, and non-viable sections of their operations.

1 Like

With repect to the poster I don't believe it. There APU line with the A10 is awsome for sub 300 builds and is a result of the synergy of being a cpu and gpu company. The fact that it is named Kavari instead of something more western like A10 Thunderbolt tell me they are heading east and EVERYTHING depends on the new markets in China and India etc.

I find this business move, assuming they go along with it, to be interesting, to say the least.

Be careful when saying that the consoles are a money maker. IIRC from an article I read a while ago, they bid for the console hardware was either at or very very marginally above cost. Consoles aren't a money maker for even Microsoft and Sony. The real money for them comes in licensing fees.

Regarding the current stock price, take the bump with a grain of salt as it's riding the hype of the new Fury/300 series reveal. The real test will be how well they actually compete against the current offerings that are out and if AMD can meet the demand. The possible de-merger makes sense as it may serve to compartmentalize any solvency issues that could come from the massive amount of debt that's coming due in 2019.

I never said the APU's were terrible. BUT as you start to ugprade the system later down the road cause the APU can't handle a certain workload, people always grab an Intel chip. why? cause they perform better. AMD hasn't catered to the mid-range and enthusiast market in years. and that is the problem. you can make cheap processors all day, but at the end of the day, when all is said and done, and you need to get work done on your machine, an APU isn't going to cut it. you're going to need a High End Chip. and you aren't going to invest into AMD's Enthusaist line when it hasn't been updated since 2012.

AMD bid at cost for the consoles. So they are not making money.

Actually IBM hold most of those firsts. But i am guessing you are talking about home use.

Just so you are aware. This isn't the 1st time AMD has considered doing this. Separating into two separate business but under one brand isn't that bad of an idea.

There is a tiny wrinkle though to do this AMD might have to get Intel's permission to do this. The reason being is that they share patents, not all of them mind you, to prevent patent trolls from suing either of the company.

My degree actual covered a lot of business administration. What AMD is considering is a viable option to help them grow their business. Having on Side of the company dedicated to making things for businesses at the server level is a smart move. We are moving toward a cloud based work space in the business sector and there will be a need to have servers to run them. Having a side dedicated to the consumer is good move. AMD has come out with some really exciting stuff for the consumer with the HBM technology. Could you imagine what would happen if they found a way to use that technology with an APU? I'm pretty excited about the possibilities.

I agree with everything you said except for one thing. Gaming rigs 300 to 450 I like the FX line today. Admittedly a niche market for sure but one I live in if I want to remain married.
Seperating consumer and business makes sense and I just never thought of that and just assumed they would seperate cpu and gpu. My bad

This isn't a new topic for AMD but a revisit. AMD would have to come up with some remarkable server chips to compete with Intel because you aren't making money in the computer industry if bushinesses aren't buying your stuff. AMD has proven you can't do well if you are just in the consumer space. Maybe splitting off the server division would allow more focus on that particular area and they can move somebody up to handle it?

they were also the first with a 1ghz graphics card too I believe.

Project Quantum kinda confirms that AMD hasn't much hope with their CPU segment. They specifically stated that it would use an i7 however there may be an option for a fully AMD version of Quantum. My guess is that by that they meant Carrizo in which case could aid the company a ton or destroy it. I'd expect the latter and that is why they are considering two subdivisions.

I believe his point was is that if you do go with an APU in a budget system you're locked to that socket, where as if you bought a decent Intel board, Pentium, and a cheap GPU, you would have a better upgrade path.

wouldn't it be might cool for enthusiast grade AMD systems to use HBM instead of DDR4?
(not really if price wasn't better than ddr4...) still having 1GB or something attached to their CPU's as L4 cache something :>

Why don't they just do that? It'd be way faster and way higher clocked

would that even help if it was mounted as L4? Base problem of their per core performance is that each 2 cores share L2 Cache, or was it L1 Cache? Main problem is their resource management ~ as today software don't use more than 4 cores and they don't distribute the data evenly. (basically a programmer configures core 1 to deal with feeding data to gpu, core 2 does something else, core 3 does ai, core 4 does something else...) this is what our programmers (cryengine does) for cores.

If AMD with their 8/12core cpu's are to compete they'll need some software/driver that will take those 4 path of info and distribute them evenly over whole span/array of cores they have, its the only way for those that do not support more than 4 cores. (those tech's are used in CF) but i believe we'll need more ram to accommodate that.

The game industry will stick to intel as main image of what they can do with cores - so we'll wait quite a while even after dx12 to see them using more than 4 cores.